O. Wolkenhauer, S. Bej, University of Rostock
Service center: RNA Bioinformatics Center – RBC
During the COVID-19 pandemic, numerous mathematical models have been developed. For most models, the focus has been on predicting the peak of new daily cases as this would give an indication of the resources that would be required from a health system. Another use of models is to demonstrate the effect of countermeasures, like social distancing. World-wide, the reporting of new infections has suffered from uncertainty due to reporting practices, and a lack of systematic and widespread testing. This has provided a major challenge for real-time predictions of case numbers and predicted peak time points.
Reporting artefacts, and the sensitivity of SEIR type models to changes in parameter values, suggests that models focussing on general pattern and relative changes in the shape of curves is a more promising approach. For the model we have developed, the aim is not an accurate prediction of case numbers in a particular population. The purpose is to predict consequences and pattern emerging from different measures, implemented to control the pandemic.
Our model is specifically focused on demonstrating the effectiveness of testing, in combination with social distancing. Using ordinary differential equations, the model will distinguish between four groups of (i) People who are 'Susceptible', who can still become infected; (ii) 'Unidentified' spreaders, that is, people who can spread the infection but have not been tested to be COVID positive yet; (iii) 'Identified' spreaders, that is, people who can spread the infection and have been tested to be COVID positive; and (iv) Resolved cases, including deaths and recovery.
We are particularly interested in "aggressive" testing, that is, testing individuals even without symptoms. To this end, we have produced a Python-based Jupyter Notebook, allowing for an interactive visualisation of scenarios that compare different implementations of testing and social distancing policies.
Since both, a total lockdown and aggressive testing will be difficult to achieve, especially in the early phase of a pandemic. Our analysis suggests however that with aggressive testing and strict isolation of identified spreaders, it is possible to achieve control of the pandemic without a total lockdown.
For further information please also read our recent publication here.